Regulation of prostatic cell growth is controlled, at least in part by steroid hormones and particularly androgens or male hormones. Epidemiological investigations have revealed that testosterone levels are relatively high in African-Americans and lower in Asians and Caucasians, a pattern that is reminiscent of the risk for developing prostate cancer among racial/ethnic groups. The normal (and abnormal) genetic variations in steroid hormone-metabolic genes in human populations are only poorly characterized and not well understood at this time. It is our hypothesis that genetic variants of human androgen- metabolic enzymes contribute significantly to androgen levels and risk for prostate cancer. We propose to identify and characterize in genetic and biochemical detail common cSNPs (coding region single nucleotide polymorphisms) in a human hormone-metabolic genes with relevance to predisposition to prostate cancer and male reproductive biology in general, the HSD17B3 gene which encodes the testicular or type III 17beta- hydroxysteroid dehydrogenase enzyme. This enzyme converts the precursor adrostenedione into testosterone, the male hormone. We propose to identify cSNPs in four large racial/ethnic population in the US: African-Americans, Asians, Caucasians and Latinos that are also at very different risk for prostate cancer. African- Americans have a very high risk for the disease, while Asians have relatively low risk. CSNPs will be identified by automated DNA sequencing of PCR (polymerase chain reaction)-amplified exonic DNA. Missense mutations will be reconstructed in the cDNA and over- expressed to assess the functional significance of these particular cSNPs. Thus, we intend to investigate the following two specific aims: 1) To identify constitutional ("germline") DNA mutations (particularly cSNPs) in a male hormone-metabolic gene, HSD17B3, in four racial/ethnic groups; and 2) to establish the in vitro biochemical properties of all missense mutations identified in 1. In summary, we will investigate the molecular genetics of natural variation in an androgen-metabolic gene in four different racial/ethnic populations. Our studies will have relevance to the understanding of predisposition to prostate cancer as well as male reproductive biology and its aberrations. The investigations proposed here will lead to a molecular epidemiologic analysis of prostate cancer risk in various racial/ethnic groups.